NVIDIA GB200 vs AMD Instinct MI250X
NVIDIA GB200
AMD Instinct MI250X
|
2x 18944 Shaders
384GB (2x 192GB) HBM3e
2112MHz
|
14080 Shaders
128GB HBM2e
1700MHz
|
|
Peak AI Performance
40 PFLOPS
FP4 Tensor Sparse
|
Peak AI Performance
382.98 TOPS
INT4 Tensor
|
|
FP32
160.04 TFLOPS
|
FP32
47.87 TFLOPS
|
|
FP16
320.08 TFLOPS
|
FP16
95.74 TFLOPS
|
|
Form Factor
Superchip
-
|
Form Factor
OAM Module
-
|
|
TDP
2700W
|
TDP
560W
|
|
-
-
-
-
-
|
-
-
-
-
-
|
|
GB6 OpenCL N/A
0%
|
GB6 OpenCL N/A
0%
|
|
GB6 Metal N/A
0%
|
GB6 Metal N/A
0%
|
|
GB6 Vulkan N/A
0%
|
GB6 Vulkan N/A
0%
|
|
GB5 OpenCL N/A
0%
|
GB5 OpenCL N/A
0%
|
|
GB5 CUDA N/A
0%
|
GB5 CUDA N/A
0%
|
|
GB5 Metal N/A
0%
|
GB5 Metal N/A
0%
|
|
GB5 Vulkan N/A
0%
|
GB5 Vulkan N/A
0%
|
|
OCT 2020.1 N/A
0%
|
OCT 2020.1 N/A
0%
|
|
OCT Metal N/A
0%
|
OCT Metal N/A
0%
|
|
Peak AI
Performance
40 PFLOPS
FP4 Tensor Sparse
|
Peak AI
Performance
382.98 TOPS
INT4 Tensor
|
|
FP4
20 PFLOPS
Tensor
40 PFLOPS
Tensor Sparse
|
-
-
-
|
|
FP8
-
10 PFLOPS
Tensor (FP16 Accumulate)
20 PFLOPS
Tensor (FP16 Accumulate) Sparse
10 PFLOPS
Tensor (FP32 Accumulate)
20 PFLOPS
Tensor (FP32 Accumulate) Sparse
|
-
-
-
-
-
-
|
|
FP16
320.08 TFLOPS
5 PFLOPS
Tensor (FP16 Accumulate)
10 PFLOPS
Tensor (FP16 Accumulate) Sparse
5 PFLOPS
Tensor (FP32 Accumulate)
10 PFLOPS
Tensor (FP32 Accumulate) Sparse
|
FP16
95.74 TFLOPS
382.98 TFLOPS
Tensor (FP16 Accumulate)
-
382.98 TFLOPS
Tensor (FP32 Accumulate)
-
|
|
FP32
160.04 TFLOPS
-
-
|
FP32
47.87 TFLOPS
95.74 TFLOPS
Tensor
-
|
|
FP64
80.02 TFLOPS
78.13 TFLOPS
Tensor
|
FP64
47.87 TFLOPS
95.74 TFLOPS
Tensor
|
|
BF16
320.08 TFLOPS
5 PFLOPS
Tensor
10 PFLOPS
Tensor Sparse
|
BF16
-
382.98 TFLOPS
Tensor
-
|
|
TF32
2.5 PFLOPS
Tensor
5 PFLOPS
Tensor Sparse
|
-
-
-
|
|
-
-
-
|
INT4
382.98 TOPS
Tensor
-
|
|
INT8
-
10 POPS
Tensor
20 POPS
Tensor Sparse
|
INT8
-
382.98 TOPS
Tensor
-
|
|
INT32
160.04 TOPS
|
-
-
|
|
-
-
|
-
-
|
|
Pixel
Fillrate
67.584 GPixel/s
|
Pixel
Fillrate
-
|
|
-
-
|
-
-
|
|
Texture
Fillrate
1250.304 GTexel/s
|
Texture
Fillrate
1496 GTexel/s
|
|
Manufacturer
NVIDIA
|
Manufacturer
AMD
|
|
Chip Designer
NVIDIA
|
Chip Designer
AMD
|
|
Architecture
Blackwell
|
Architecture
CDNA 2
|
|
Family
Server
|
Family
Instinct
|
|
Codename
NV190
GB100
Variant
Oberon
|
Codename
Aldebaran
Aldebaran XTX
Variant
Aldebaran XTX
|
|
Market Segment
Server
|
Market Segment
Server
|
|
Release Date
3/18/2024
|
Release Date
11/8/2021
|
|
Foundry
TSMC
-
|
Foundry
TSMC
-
|
|
Fabrication Node
4NP
-
|
Fabrication Node
N6
-
|
|
Die Size
4x 810 mm²
-
|
Die Size
2x 724 mm²
-
|
|
Transistor Count
4x 104 Billion
-
|
Transistor Count
2x 28 Billion
-
|
|
Transistor Density
128.40M/mm²
-
|
Transistor Density
38.67M/mm²
-
|
|
Form
Superchip
|
Form
OAM Module
|
|
Shading Units
2x 18944 Shaders
-
|
Shading Units
14080 Shaders
-
|
|
Texture Mapping Units
2x 592 TMUs
|
Texture Mapping Units
880 TMUs
|
|
Render Output Units
2x 32 ROPs
|
Render Output Units
-
|
|
Tensor Cores
2x 592 T-Cores
|
Tensor Cores
880 T-Cores
|
|
-
-
|
-
-
|
|
Streaming Multiprocessors
2x 148 SMs
|
-
-
|
|
-
-
|
Compute Units
220 CUs
|
|
-
-
|
-
-
|
|
-
-
|
-
-
|
|
2062MHz Tensor
-
-
2112MHz
|
-
-
1000MHz Base
1700MHz
|
|
-
-
|
-
-
|
|
L1
64KB/SM Tex
256KB/SM
-
-
|
L1
-
-
16KB/CU
-
|
|
L2
64MB Shared
|
L2
16MB Shared
|
|
-
-
-
|
-
-
-
|
|
384GB (2x 192GB)
HBM3e
-
|
128GB
HBM2e
ECC
|
|
Bus Width
8192Bit
|
Bus Width
8192Bit
|
|
Clock
3906MHz
Transfer Rate
7.8GT/s
Bandwidth
8000.1GB/s
|
Clock
1600MHz
Transfer Rate
3.2GT/s
Bandwidth
3276.8GB/s
|
|
-
-
-
-
-
-
-
-
-
|
-
-
-
-
-
-
-
-
-
|
|
TDP
2700W
|
TDP
560W
|
|
-
-
|
-
-
|
|
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
No Ports
|
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
No Ports
|
|
Max Resolution
Unknown
|
Max Resolution
Unknown
|
|
Max Resolution Refresh Rate
-
|
Max Resolution Refresh Rate
-
|
|
Variable Refresh Rate
G-Sync
FreeSync
-
|
Variable Refresh Rate
-
-
-
|
|
Display Stream Compression (DSC)
Not Supported
|
Display Stream Compression (DSC)
Not Supported
|
|
Multi Monitor Support
Unknown
|
Multi Monitor Support
Unknown
|
|
-
-
|
-
-
|
|
No Encoders
-
|
Model
2x VCN 2.6
|
|
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
|
Codec
-
-
-
-
-
-
-
-
AVC (H.264)
HEVC (H.265)
-
-
-
-
|
|
Model
7x NVDEC 6
|
No Decoders
|
|
Codec
MPEG-1
MPEG-2
MPEG-4
-
VC-1
VP8
VP9
-
AVC (H.264)
HEVC (H.265)
-
AV1
-
-
|
-
-
-
-
-
-
-
-
-
-
-
-
-
-
|
|
-
-
-
-
|
-
-
-
-
|
|
-
-
OpenCL
3.0
-
-
|
-
-
OpenCL
3.0
-
-
|
|
-
-
CUDA
10.0
-
-
PureVideo HD
VP13
VDPAU
Feature Set M
|
-
-
-
-
GFX
9.4
-
-
-
-
|
|
Not a Card
-
-
-
|
-
-
-
-
|
|
-
-
-
-
-
-
-
-
|
-
-
-
-
-
-
-
-
|
|
-
-
PCIe Version
6.0
PCIe Lanes
16
|
-
-
PCIe Version
4.0
PCIe Lanes
16
|
|
Multi GPU Support
Supported
Type
NVLink
|
Multi GPU Support
Supported
Type
Infinity Fabric
|
|
-
-
-
-
-
-
|
-
-
-
-
-
-
|
Copy Link